CN106682788A - Method and device for selection sequence determination - Google Patents
Method and device for selection sequence determination Download PDFInfo
- Publication number
- CN106682788A CN106682788A CN201710012887.XA CN201710012887A CN106682788A CN 106682788 A CN106682788 A CN 106682788A CN 201710012887 A CN201710012887 A CN 201710012887A CN 106682788 A CN106682788 A CN 106682788A
- Authority
- CN
- China
- Prior art keywords
- particle
- filial generation
- fitness
- population
- parent population
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
Abstract
The present invention discloses a method and device for selection sequence determination. The method comprises: establishing a target function of a selection sequence problem according an optimal object, employing the particle swarm optimization to initialize the target function, generating a parent particle swarm, performing fitness comparison, determining the individual extreme value corresponding to each particle and the local extreme value of the parent particle swarm. According to the individual extreme value and the local extreme value, performing crossover and variation operation of the parent particle swarm, generating a filial generation particle swarm, taking the filial generation particle swarm as a parent particle swarm of the next iteration, repeating the execution of the operations mentioned above, when the number of times of the iteration reaches the preset maximum number of times of the iteration, stopping the repeat execution of the operations, and outputting the result of the selection sequence problem. According to the technical scheme, the early-maturing problem caused by the particle swarm optimization approaching to the optimal solution can be effectively avoided so as to effectively avoid the condition that the optimal solution is the local optimum and determine a selection sequence with the least time consumption.
Description
Technical field
The present invention relates to warehousing management technical field, more particularly to a kind of selection based on genetic idea particle cluster algorithm
The method and device that order determines.
Background technology
Warehousing management refers to and the goods and materials of the storage in warehouse and warehouse is managed, due to modern storage effect not only
It is keeping, is more flow of material center, the safety of materials storage is not exclusively conceived to the emphasis of warehousing management, more
How it is concerned with modern technologies, such as information technology, automatic technology is improving the speed and benefit of storage running more.
The substance of warehousing management includes Delivery and input work of goods etc., and lane stacker is warehousing management
In commonly use instrument, can complete goods in the range of shelf vertical and horizontal movement, will be located at access adit goods be stored in
Goods handling in goods lattice, or taking-up goods lattice realizes the 3 D stereo selection of goods to access adit.Goods in fixed goods shelf
Goods yard point is relatively fixed, and using lane stacker goods selection in shelf is fixed, and generally, chooses the quantity of goods
More, the selection time that different selection orders are spent is often different, if being capable of determining that, selection time minimum one kind is picked
Choosing order, will greatly save the time that goods spends of choosing, so as to improve the work efficiency of selection.
In traditional approach, selection order is optimized using particle cluster algorithm, to expect to determine that optimal solution is i.e. optimum
Selection order, according to selection order so that the time that lane stacker selection goods spends is minimum.But particle cluster algorithm
Initial stage convergence rate is very fast, and when the close optimal solution of particle, speed almost vanishing causes convergence rate slack-off so that grain
There is homoplasy in subgroup, brings precocity, namely the optimal solution determined the situation of local optimum occur.
Therefore, how to improve the premature convergence problem of particle cluster algorithm, be this area skill so that it is determined that going out the selection order of optimum
Art personnel technical problem urgently to be resolved hurrily.
The content of the invention
It is an object of the invention to provide the method and dress of a kind of selection order determination based on genetic idea particle cluster algorithm
Put, the premature convergence problem of particle cluster algorithm can be effectively improved, so that it is determined that the selection order for going out a kind of optimum spends the time most
Few selection order.
To solve above-mentioned technical problem, the present invention provides a kind of selection order based on genetic idea particle cluster algorithm and determines
Method include：
S1, according to optimization aim, set up the object function of selection sequencing problem；
S2, object function is initialized using particle cluster algorithm, generate parent population；
S3, the fitness for calculating each particle in the parent population；
The personal best particle of the fitness of intended particle and the intended particle in parent population described in S4, comparison
Fitness, using fitness optimal value as the intended particle individual extreme value；The intended particle is the parent population
In any one particle；
Colony's optimal location of the fitness of each particle and the parent population in parent population described in S5, comparison
Fitness, using fitness optimal value as the parent population local extremum；
S6, according to the individual extreme value and the local extremum, cross and variation operation is carried out to the parent population, it is raw
Into filial generation population；
S7, using the filial generation population as next iteration parent population；
S8, judge whether iterationses reach default maximum iteration time, if so, the knot of output selection sequencing problem
Really；If otherwise returning the S3.
Optionally, the object function is specially：
Wherein, T represents the time that selection order spends, and n represents goods yard point number, xiRepresent the row number of goods yard point i, ziTable
Show the line number of goods yard point i, b is goods lattice width, and h is goods lattice height；vxFor horizontal movement speed, vzFor the speed that moves vertically.
Optionally, in the S6：
S61：The intended particle individual extreme value corresponding with the intended particle is carried out into crossover operation, first is produced
Filial generation particle, the fitness of the comparison intended particle and the fitness of the F1 particle；By the grain that fitness is optimum
Son produces second filial generation population as second filial generation particle；The second filial generation population is sub by described at least one second
Constitute for particle；
S62, the second filial generation particle and the local extremum are carried out crossover operation, produce F3 particle, than
The fitness of the fitness of the second filial generation particle and the F3 particle；Using the optimum particle of fitness as the
Four filial generation particles, produce the 4th filial generation population；The 4th filial generation population is by described at least one the 4th filial generation particle structures
Into；
S63, the 4th filial generation particle is carried out mutation operation, produce the 5th filial generation particle, comparison the 4th filial generation
The fitness of the fitness of particle and the 5th filial generation particle；Using the optimum population of fitness as the 6th filial generation particle,
Produce the 6th filial generation population；The 6th filial generation population is made up of described at least one the 6th filial generation particles；
It is described using the filial generation population as next iteration parent population, including：By the 6th filial generation grain
Parent population of the subgroup as next iteration.
Optionally, in the S61：
The intended particle individual extreme value corresponding with the intended particle is carried out using partial mapped crossover operator
Crossover operation, produces F1 particle.
Optionally, in the S62：
The second filial generation particle is carried out into crossover operation with the local extremum using partial mapped crossover operator, is produced
F3 particle.
Optionally, in the S63：
The 4th filial generation particle is carried out into mutation operation using sequence variation mode is changed, the 5th filial generation particle is produced.
The present invention also provides a kind of selection order determining device based on genetic idea particle cluster algorithm to be included, the selection
Order determining device includes setting up unit, signal generating unit, computing unit, comparing unit, cross and variation unit, replacement unit and sentencing
Disconnected unit：
It is described to set up unit, for according to optimization aim, setting up the object function of selection sequencing problem；
The signal generating unit, for initializing to object function using particle cluster algorithm, generates parent population；
The computing unit, for calculating the parent population in each particle fitness；
The comparing unit, the fitness and the intended particle for intended particle in relatively more described parent population
The fitness of personal best particle, using fitness optimal value as the intended particle individual extreme value；The intended particle is
Any one particle in the parent population；
The comparing unit is additionally operable to the fitness of each particle and the parent particle in parent population described in comparison
Group colony's optimal location fitness, using fitness optimal value as the parent population local extremum；
The cross and variation unit, for according to the individual extreme value and the local extremum, to the parent population
Cross and variation operation is carried out, filial generation population is generated；
The replacement unit, for using the filial generation population as next iteration parent population；
The judging unit, for judging whether iterationses reach default maximum iteration time, if so, output selection
The result of sequencing problem；If otherwise triggering the computing unit.
Optionally, the object function is specially：
Wherein, T represents the time that selection order spends, and n represents goods yard point number, xiRepresent the row number of goods yard point i, ziTable
Show the line number of goods yard point i, b is goods lattice width, and h is goods lattice height；vxFor horizontal movement speed, vzFor the speed that moves vertically.
Optionally, the cross and variation unit includes intersecting subelement and variation subelement：
The intersection subelement, for the intended particle individual extreme value corresponding with the intended particle to be handed over
Fork operation, produces F1 particle, the fitness of the comparison intended particle and the fitness of the F1 particle；Will
The optimum particle of fitness produces second filial generation population as second filial generation particle；The second filial generation population is by described
At least one second filial generation particle is constituted；
The intersection subelement is additionally operable to for the second filial generation particle and the local extremum to carry out crossover operation, produces
F3 particle, the fitness of the comparison second filial generation particle and the fitness of the F3 particle；By fitness
Optimum particle produces the 4th filial generation population as the 4th filial generation particle；The 4th filial generation population is by described at least one
Individual 4th filial generation particle is constituted；
The variation subelement, for the 4th filial generation particle to be carried out into mutation operation, produces the 5th filial generation particle, than
The fitness of the 4th filial generation particle and the fitness of the 5th filial generation particle；Using the optimum population of fitness as
6th filial generation particle, produces the 6th filial generation population；The 6th filial generation population is by described at least one the 6th filial generation particles
Constitute；
The replacement unit specifically for using the 6th filial generation population as next iteration parent population.
By above-mentioned technical proposal as can be seen that according to optimization aim, the object function of selection sequencing problem is set up, using grain
Swarm optimization is initialized to object function, parent population is generated, by the way that each particle in parent population is right with it
The personal best particle answered carries out the comparison of fitness, determines the corresponding individual extreme value of each particle, and by parent population
In each particle and colony's optimal location carry out the comparison of fitness, determine the local extremum of parent population.By foundation
The individual extreme value and the local extremum, cross and variation operation is carried out to the parent population, generates filial generation population, will
The filial generation population repeats aforesaid operations as the parent population of next iteration, until iterationses reach it is pre-
If maximum iteration time, then stop repeating aforesaid operations, the result of output selection sequencing problem.Above-mentioned technical proposal can
With the premature convergence problem for effectively avoiding being brought during the close optimal solution of particle cluster algorithm, so as to be prevented effectively from optimal solution for local optimum
Situation, may thereby determine that out a kind of cost time it is minimum selection order.
Description of the drawings
In order to be illustrated more clearly that the embodiment of the present invention, the accompanying drawing to be used needed for embodiment will be done simply below
Introduce, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for ordinary skill people
For member, on the premise of not paying creative work, can be with according to these other accompanying drawings of accompanying drawings acquisition.
Fig. 1 is that a kind of selection order based on genetic idea particle cluster algorithm provided in an embodiment of the present invention determines method
Flow chart；
Fig. 2 is a kind of flow chart of cross and variation operational approach provided in an embodiment of the present invention；
Fig. 3 is a kind of selection order determining device based on genetic idea particle cluster algorithm provided in an embodiment of the present invention
Structure drawing of device.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Based on this
Embodiment in invention, those of ordinary skill in the art are not under the premise of creative work is made, and what is obtained is every other
Embodiment, belongs to the scope of the present invention.
In order that those skilled in the art more fully understand the present invention program, with reference to the accompanying drawings and detailed description
The present invention is described in further detail.
Next, it is suitable that a kind of selection based on genetic idea particle cluster algorithm that the embodiment of the present invention provided is discussed in detail
Sequence determines method.Method provided in an embodiment of the present invention, when can apply to choose goods, determines a kind of selection cost most
Little selection order.The time that selection cost spends when can be specifically and choose goods according to selection order.
Fig. 1 is that a kind of selection order based on genetic idea particle cluster algorithm provided in an embodiment of the present invention determines method
Flow chart, the selection order determination method includes：
S1, according to optimization aim, set up the object function of selection sequencing problem.
In actual applications, when being chosen to goods using lane stacker, each or every class goods have what it was fixed
Deposit position is goods yard point, and in the embodiment of the present invention, selection order can be the order for choosing the goods yard point passed through during goods.
In the embodiment of the present invention, for example, can there are 9 goods yard points, Ke Yiyong to represent selection order by the way of sequential encoding
This 9 unduplicated natural numbers of 1-9 are representing.
Different selection orders, its time for spending is not quite similar.Optimization aim can be to determine out a kind of cost time
Minimum selection order.
Corresponding object function can be set up according to optimization aim, the object function can be that by the optimization mesh
One function of target.Because optimization aim is minimum with regard to the cost time, the object function for hence setting up can be one and when
Between related function, the object function is specifically as follows：
Wherein, T represents the time that selection order spends, and n represents goods yard point number, xiRepresent the row number of goods yard point i, ziTable
Show the line number of goods yard point i, xi+1Represent the row number of goods yard point i+1, zi+1The line number of goods yard point i+1 is represented, b is goods lattice width, and h is
Goods lattice height；vxFor horizontal movement speed, vzFor the speed that moves vertically.
In embodiments of the present invention, the time that selection order spends can be lane stacker from origin such as access adit
Set out and travel through each goods yard point once and only once, complete selection goods task and eventually pass back to the time that origin is used.
S2, object function is initialized using particle cluster algorithm, generate parent population.
Parent population can be the population for initially generating during an iteration.
Object function is initialized using particle cluster algorithm, can at random generate m particle in n dimension solution spaces,
In embodiments of the present invention, can be using the m particle for generating as parent population, each particle in parent population has it
Corresponding position and speed.
N dimension solution spaces can be the two-dimensional matrix of a n row m row, and each particle takes a line.The concrete value of n can be with
It is the goods yard point number for choosing goods.The embodiment of the present invention, the concrete numerical value for m is not limited, and can be according to optimization
Specific requirement determines, or with reference to the value determination of n, the value of such as m can be the multiple of n values.
S3, the fitness for calculating each particle in the parent population.
Each particle can represent a kind of selection order, and the fitness of particle can reflect spends time taking length.
The fitness of each particle can be calculated according to fitness function, in embodiments of the present invention, fitness function can
Being can be with object function identical function, i.e. fitness function：
In embodiments of the present invention, it is thus necessary to determine that go out the cost time it is minimum selection order, so, the adaptation of each particle
Degree is less, illustrates that the time that the selection order that the particle is represented spends is fewer.
The personal best particle of the fitness of intended particle and the intended particle in parent population described in S4, comparison
Fitness, using fitness optimal value as the intended particle individual extreme value；The intended particle is the parent population
In any one particle.
Each particle has its corresponding individual extreme value, individual extreme value can the particle obtain in optimal location
Fitness value, i.e., the current optimal solution for finding.
Using particle cluster algorithm, when generating parent population, each particle can have its corresponding individual optimum position
Put, m particle randomly generated in above process, then this m particle its corresponding have m personal best particle.In initialization
When, the fitness value of the corresponding personal best particle of particle is defaulted as into the corresponding individual extreme value of the particle.
In order to the corresponding individual extreme value of the particle further clearly given tacit consent to be whether the particle optimal location obtain it is suitable
Angle value is answered, the fitness of the corresponding personal best particle of the fitness of each particle can be compared, so as to choose
Go out the corresponding individual extreme value of each particle.
By taking a particle as an example, when the fitness of the particle is less than the fitness of its corresponding personal best particle, then
Can using the fitness value of the particle as the particle individual extreme value；In the same manner, when the fitness of the particle is corresponding more than its
During the fitness of personal best particle, then can the personal best particle fitness value as the particle individual pole
Value.When the fitness of the particle is equal to the fitness of its corresponding personal best particle, then can choose therein
Any one as individual extreme value, for example, it may be using the fitness value of the particle as the particle individual extreme value, or
Will the personal best particle fitness value as the particle individual extreme value.
Colony's optimal location of the fitness of each particle and the parent population in parent population described in S5, comparison
Fitness, using fitness optimal value as the parent population local extremum.
Its corresponding local extremum is understood in a population, local extremum can be optimal location in the population
Particle fitness value, i.e., in the population find globally optimal solution.
Using particle cluster algorithm, when generating parent population, the parent population has corresponding colony optimum position
Put.In initialization, the fitness value of colony's optimal location of population is defaulted as into the corresponding local extremum of the population.
In order to whether the corresponding local extremum of the population further clearly given tacit consent to is the grain of optimal location in the population
The fitness value of son, can be compared the fitness of each particle with the fitness of colony's optimal location, so as to choose
Go out the corresponding local extremum of the population.
In the embodiment of the present invention, for the mode of the fitness of the fitness and colony's optimal location for comparing each particle
Do not limit, each particle in the population can be carried out successively the comparison of fitness with colony's optimal location, finally
The local extremum of the population is determined, for example, 30 particles, respectively particle 1- particles 30 is included in a population,
The population is compared particle 1 with the fitness of colony's optimal location to there is colony's optimal location, selects
Fitness is less as new colony's optimal location, then the fitness of the colony optimal location new with this of particle 2 is compared
Compared with, it is less as new colony's optimal location to select fitness, the like, until by all of particle ratio of the population
Relatively finish, then the fitness value of the new colony's optimal location for finally giving is the local extremum of the population.
Can also first be compared the fitness of each particle in the population, select the minimum grain of fitness
Son, then the particle and colony's optimal location are carried out into the comparison of fitness, if the fitness of the particle is optimum less than the colony
The fitness value of position, then using the fitness value of the particle as the population local extremum, if the fitness of the particle is big
In the fitness value of colony's optimal location, then can colony's optimal location fitness value as the population office
Portion's extreme value.When the fitness of the particle is equal to the fitness value of colony's optimal location, then can choose therein
Any one as local extremum, for example, it may be using the fitness value of the particle as the population local extremum, or
Be will the personal best particle fitness value as the population local extremum.
By the way that the fitness value of each particle in particle colony and the fitness value of colony's optimal location are compared,
Can determine that the local extremum of the fitness value as the particle colony of an optimum.
S6, according to the individual extreme value and the local extremum, cross and variation operation is carried out to the parent population, it is raw
Into filial generation population.
When being optimized using particle cluster algorithm, when the close optimal location of particle, the optimal location can be used for mark
The cost time it is minimum a kind of selection order, convergence rate can be slack-off, causes population homoplasy occur, brings precocity, in order to
The impact that the precocity is caused is reduced, in the embodiment of the present invention, the cross and variation operation in genetic algorithm can be adopted, to the particle
Group is processed, and generates filial generation population.Compared to parent population, the close optimal location of particle in the filial generation population
When, generation locally optimal solution can be prevented effectively from.
With a population P={ p1,p2,...,pNAs a example by, wherein p represents single particle, and N represents population size, i.e. grain
The number of the particle included in subgroup.Each particle by the position that is located and speed describing, the particle of diverse location its fit
Answer angle value different.Particle piThe positional representation for tieing up solution space in d is Xi=[xi1,xi2,...,xid]T, and speed is expressed as Vi=
[vi1,vi2,...,vid]T.Particle piIndividual extrema representation be pbesti, i.e. pbestiIt is particle piThe current optimal solution for finding.
piLocal extremum be expressed as gbesti, i.e. gbestiIt is particle piThe globally optimal solution of place particle group discovery.
Assume that i-th particle is X in the position in the n-th generation and speedi (n)And Vi (n), then the information of the (n+1)th generation particle be:
Vi (n+1)=wVi (n)+c1·rand()·(pbesti-Xi (n))+c2·rand()·(gbesti-Xi (n)) (4)
Wherein, w is a normal number, referred to as inertia weight coefficient, c1、c2Respectively represent particle self-acceleration weight and
Global acceleration weight, rand () is that equally distributed random number is obeyed in [0,1].
The particle cluster algorithm for incorporating genetic idea is by the wV in formula (4)i (n)Item regards a kind of mutation operation as,
c1·rand()·(pbesti-Xi (n))+c2·rand()·(gbesti-Xi (n)) item regards a kind of crossover operation as.
S7, using the filial generation population as next iteration parent population.
S8, judge whether iterationses reach default maximum iteration time, if so, the knot of output selection sequencing problem
Really；If otherwise returning the S3.
Include m particle in population in aforesaid operations, each particle can be used for identifying a kind of selection order, lead to
Cross and be constantly optimized the position of particle, will the continuous close optimal location of the particle.By iterative operation, in the population
Each particle be eventually intended to same optimal location, that is, choose the result of sequencing problem.
Default maximum iteration time, can be configured according to the demand of practical operation.When iterationses reach it is default
Maximum iteration time after, the result of selection sequencing problem can be exported, the result as identifies cost time minimum one kind
Selection order.
It should be noted that have its corresponding parent population and filial generation population in iterative process each time, often
Parent population and the parent population in last iterative process can be with different in secondary iterative process, and each iterative process is obtained
The filial generation population that obtains from last iterative process of filial generation population can be with different.
Next, the specific operation process that cross and variation operation is carried out to the parent population is launched to introduce, such as Fig. 2 institutes
Show, be a kind of mode of feasible cross and variation operation, concrete operations are as follows：
S61：The intended particle individual extreme value corresponding with the intended particle is carried out into crossover operation, first is produced
Filial generation particle, the fitness of the comparison intended particle and the fitness of the F1 particle；By the grain that fitness is optimum
Son produces second filial generation population as second filial generation particle；The second filial generation population is sub by described at least one second
Constitute for particle.
In embodiments of the present invention, each particle can be used for identifying a kind of selection order, and the selection order can pass through
One coded strings represents this 9 goods yard points, coded strings representing for example, there is 9 goods yard points with natural number 1-9 respectively
123456789 represent a kind of selection order from goods yard point 1 successively arrival site 9.
The intended particle individual extreme value corresponding with the intended particle is carried out into crossover operation, can be specifically by the mesh
The corresponding coded strings of mark particle coded strings corresponding with the individual extreme value are using partial mapped crossover (Partially Mapped
Crossover, PMX) carry out crossover operation.For example
| 3456 | of A=1 2789
| 7654 | of B=9 8321
Wherein, A can be the corresponding coded strings of a particle in parent population, can be referred to as parent coded strings, and B can
Being the corresponding coded strings of individual extreme value corresponding with the particle.
Two cross points, such as X=3, Y=6, according to corresponding mapping relations are randomly choosed first in parent coded strings
The gene string of 3 to 6 in A, B is exchanged, then A, B is repaired, if the gene in intersection region has with the gene not exchanged
Repeat, be then replaced according to the mapping relations of gene string.The mapping relations of this example are：
Mapping relations：3-7,4-6,5-5
Calculated using PMX after terminating, can obtain newly organized sequence is：
| 7654 | of A'=1 2389
| 3456 | of B'=9 8721
Wherein, the particle corresponding to coded strings that A' is represented is the F1 particle for obtaining, by the F1 grain
The fitness of the particle corresponding to coded strings that son is represented with A is compared, when the fitness of the particle is less than F1 grain
During the fitness of son, then can be using the particle as second filial generation particle；When the fitness of the particle is more than F1 particle
Fitness when, then can be using the F1 particle as second filial generation particle.For the fitness of the particle is equal to first
The situation of the fitness of filial generation particle, then can choose it is therein any one as second filial generation particle.
By that analogy, the fitness of the corresponding F1 particle of each particle can be compared, selects suitable
The less particle of response as second filial generation particle, so as to complete the renewal to particle position so that each grain in population
Son can be more nearly optimal location.
The corresponding individual extreme value of each particle in parent population can be obtained in above-mentioned S4, by each particle and the grain
The corresponding individual extreme value of son is calculated using partial mapped crossover, F1 particle can be produced, in parent population
In include m particle, it is corresponding to produce m F1 particle, by the corresponding F1 grain of each particle
The fitness of son is compared, and selects the less particle of fitness as second filial generation particle, corresponding to produce m second
Filial generation particle, the m second filial generation particle may be constructed a second filial generation population.
Wherein the calculation of the fitness of particle is similar with the calculation of the fitness that particle is calculated in above-mentioned S3,
This is repeated no more.
S62, the second filial generation particle and the local extremum are carried out crossover operation, produce F3 particle, than
The fitness of the fitness of the second filial generation particle and the F3 particle；Using the optimum particle of fitness as the
Four filial generation particles, produce the 4th filial generation population；The 4th filial generation population is by described at least one the 4th filial generation particle structures
Into.
Crossover operation in the step is similar with the mode of crossover operation in above-mentioned S61, will not be described here.
The adaptation of the local extremum by will obtain in each second filial generation particle in second filial generation population and above-mentioned S5
Degree carries out crossover operation, produces F3 particle.Each second filial generation particle has its correspondence in second filial generation population
A F3 particle.
By the way that second filial generation particle is compared with the fitness of F3 particle, the less particle of fitness is selected
As the 4th filial generation particle, so as to complete the renewal to particle position so that each particle in population can more adjunction
Nearly optimal location.
Wherein the calculation of the fitness of particle is similar with the calculation of the fitness that particle is calculated in above-mentioned S3,
This is repeated no more.
M second filial generation particle can be obtained in above-mentioned S61, a second filial generation population is may be constructed.By S62
Operation the position of m second filial generation particle can further be optimized, obtain m the 4th filial generation particles, can be with structure
Into a 4th filial generation population.Compared with second filial generation population, the particle in the 4th filial generation population is more nearly most
Excellent position.
S63, the 4th filial generation particle is carried out mutation operation, produce the 5th filial generation particle, comparison the 4th filial generation
The fitness of the fitness of particle and the 5th filial generation particle；Using the optimum population of fitness as the 6th filial generation particle,
Produce the 6th filial generation population；The 6th filial generation population is made up of described at least one the 6th filial generation particles.
In the embodiment of the present invention can with using change sequence variation by the way of carry out mutation operation, select two in coded strings at random
Individual position, is inverted numeral therebetween, can thus avoid the generation of illegal solution.It is illustrated below：
Assume that the coded strings for having a particle are：
L=1 35724689
Wherein, L can be the corresponding coded strings of a particle in the 4th filial generation population.
The 3rd and the 5th in random selection coded strings carries out changing sequence variation, and new coded strings are：
N=1 32754689
Wherein, the particle corresponding to coded strings that N is represented is the 5th filial generation particle for obtaining.
By the way that the 4th filial generation particle and the fitness of the 5th filial generation particle are compared, the less particle of fitness is selected
As the 6th filial generation particle, so as to complete the renewal to particle position so that each particle in population can more adjunction
Nearly optimal location.
Wherein the calculation of the fitness of particle is similar with the calculation of the fitness that particle is calculated in above-mentioned S3,
This is repeated no more.
The position of m the 4th filial generation particle can further be optimized by the operation of S63, be obtained m the 6th
Filial generation particle, may be constructed a 6th filial generation population.Compared with the 4th filial generation population, in the 6th filial generation population
Particle is more nearly optimal location.
In the embodiment of the present invention, the execution sequence for S63 and S61-S62 is not limited, can be first carry out S61 and
S62, then perform S63, or first carry out S63, then perform S61 and S62.
Said process S61-S63 is the concrete introduction to cross and variation operation in S6, thus produces the 6th filial generation population
Filial generation population as described in S6, so, include the filial generation population as the parent population of next iteration：
Using the 6th filial generation population as next iteration parent population.
By above-mentioned technical proposal as can be seen that according to optimization aim, the object function of selection sequencing problem is set up, using grain
Swarm optimization is initialized to object function, parent population is generated, by the way that each particle in parent population is right with it
The personal best particle answered carries out the comparison of fitness, determines the corresponding individual extreme value of each particle, and by parent population
In each particle and colony's optimal location carry out the comparison of fitness, determine the local extremum of parent population.By foundation
The individual extreme value and the local extremum, cross and variation operation is carried out to the parent population, generates filial generation population, will
The filial generation population repeats aforesaid operations as the parent population of next iteration, until iterationses reach it is pre-
If maximum iteration time, then stop repeating aforesaid operations, the result of output selection sequencing problem.Above-mentioned technical proposal can
With the premature convergence problem for effectively avoiding being brought during the close optimal solution of particle cluster algorithm, so as to be prevented effectively from optimal solution for local optimum
Situation, may thereby determine that out a kind of cost time it is minimum selection order.
Fig. 3 is a kind of selection order determining device based on genetic idea particle cluster algorithm provided in an embodiment of the present invention
Structure drawing of device, the selection order determining device includes setting up unit 31, signal generating unit 32, computing unit 33, comparing unit
34th, cross and variation unit 35, replacement unit 36 and judging unit 37：
It is described to set up unit 31, for according to optimization aim, setting up the object function of selection sequencing problem.
The signal generating unit 32, for initializing to object function using particle cluster algorithm, generates parent population.
The computing unit 33, for calculating the parent population in each particle fitness.
The comparing unit 34, for the fitness and the intended particle of intended particle in relatively more described parent population
Personal best particle fitness, using fitness optimal value as the intended particle individual extreme value；The intended particle
For any one particle in the parent population.
The comparing unit 34 is additionally operable to the fitness of each particle and the parent grain in parent population described in comparison
The fitness of colony's optimal location of subgroup, using fitness optimal value as the parent population local extremum.
The cross and variation unit 35, for according to the individual extreme value and the local extremum, to the parent particle
Group carries out cross and variation operation, generates filial generation population.
The replacement unit 36, for using the filial generation population as next iteration parent population.
The judging unit 37, for judging whether iterationses reach default maximum iteration time, if so, output is picked
Select the result of sequencing problem；If otherwise triggering the computing unit.
Optionally, the object function is specially：
Wherein, T represents the time that selection order spends, and n represents goods yard point number, xiRepresent the row number of goods yard point i, ziTable
Show the line number of goods yard point i, xi+1Represent the row number of goods yard point i+1, zi+1The line number of goods yard point i+1 is represented, b is goods lattice width, and h is
Goods lattice height；vxFor horizontal movement speed, vzFor the speed that moves vertically.
Optionally, the cross and variation unit includes intersecting subelement and variation subelement：
The intersection subelement, for the intended particle individual extreme value corresponding with the intended particle to be handed over
Fork operation, produces F1 particle, the fitness of the comparison intended particle and the fitness of the F1 particle；Will
The optimum particle of fitness produces second filial generation population as second filial generation particle；The second filial generation population is by described
At least one second filial generation particle is constituted；
The intersection subelement is additionally operable to for the second filial generation particle and the local extremum to carry out crossover operation, produces
F3 particle, the fitness of the comparison second filial generation particle and the fitness of the F3 particle；By fitness
Optimum particle produces the 4th filial generation population as the 4th filial generation particle；The 4th filial generation population is by described at least one
Individual 4th filial generation particle is constituted；
The variation subelement, for the 4th filial generation particle to be carried out into mutation operation, produces the 5th filial generation particle, than
The fitness of the 4th filial generation particle and the fitness of the 5th filial generation particle；Using the optimum population of fitness as
6th filial generation particle, produces the 6th filial generation population；The 6th filial generation population is by described at least one the 6th filial generation particles
Constitute；
The replacement unit specifically for using the 6th filial generation population as next iteration parent population.
The explanation of feature in embodiment corresponding to Fig. 3 may refer to the related description of embodiment corresponding to Fig. 1, Fig. 2, here
No longer repeat one by one.
By above-mentioned technical proposal as can be seen that according to optimization aim, the object function of selection sequencing problem is set up, using grain
Swarm optimization is initialized to object function, parent population is generated, by the way that each particle in parent population is right with it
The personal best particle answered carries out the comparison of fitness, determines the corresponding individual extreme value of each particle, and by parent population
In each particle and colony's optimal location carry out the comparison of fitness, determine the local extremum of parent population.By foundation
The individual extreme value and the local extremum, cross and variation operation is carried out to the parent population, generates filial generation population, will
The filial generation population repeats aforesaid operations as the parent population of next iteration, until iterationses reach it is pre-
If maximum iteration time, then stop repeating aforesaid operations, the result of output selection sequencing problem.Above-mentioned technical proposal can
With the premature convergence problem for effectively avoiding being brought during the close optimal solution of particle cluster algorithm, so as to be prevented effectively from optimal solution for local optimum
Situation, may thereby determine that out a kind of cost time it is minimum selection order.
Above method and device is entered to be determined to the selection order based on genetic idea particle cluster algorithm provided by the present invention
Go and be discussed in detail.Each embodiment is described by the way of progressive in description, what each embodiment was stressed be with
The difference of other embodiment, between each embodiment identical similar portion mutually referring to.For disclosed in embodiment
For device, because it is corresponded to the method disclosed in Example, so description is fairly simple, related part is referring to method portion
Defend oneself bright.It should be pointed out that for those skilled in the art, in the premise without departing from the principle of the invention
Under, some improvement and modification can also be carried out to the present invention, these are improved and modification also falls into the protection of the claims in the present invention
In the range of.
Professional further appreciates that, with reference to the unit of each example of the embodiments described herein description
And algorithm steps, can with electronic hardware, computer software or the two be implemented in combination in, in order to clearly demonstrate hardware and
The interchangeability of software, according to function has generally described the composition and step of each example in the above description.These
Function is performed with hardware or software mode actually, depending on the application-specific and design constraint of technical scheme.Specialty
Technical staff can use different methods to realize described function to each specific application, but this realization should not
Think beyond the scope of this invention.
The step of method described with reference to the embodiments described herein or algorithm, directly can be held with hardware, processor
Capable software module, or the combination of the two is implementing.Software module can be placed in random access memory (RAM), internal memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, depositor, hard disk, moveable magnetic disc, CD-ROM or technology
In field in known any other form of storage medium.
Claims (9)
1. a kind of selection order based on genetic idea particle cluster algorithm determines method, it is characterised in that the selection order is really
The method of determining includes：
S1, according to optimization aim, set up the object function of selection sequencing problem；
S2, object function is initialized using particle cluster algorithm, generate parent population；
S3, the fitness for calculating each particle in the parent population；
The adaptation of the personal best particle of the fitness of intended particle and the intended particle in parent population described in S4, comparison
Degree, using fitness optimal value as the intended particle individual extreme value；The intended particle is in the parent population
Any one particle；
Colony's optimal location of the fitness of each particle and the parent population is suitable in parent population described in S5, comparison
Response, using fitness optimal value as the parent population local extremum；
S6, according to the individual extreme value and the local extremum, carry out cross and variation operation to the parent population, generate son
For population；
S7, using the filial generation population as next iteration parent population；
S8, judge whether iterationses reach default maximum iteration time, if so, the result of output selection sequencing problem；If
Otherwise return the S3.
2. selection order according to claim 1 determines method, it is characterised in that the object function is specially：
Wherein, T represents the time that selection order spends, and n represents goods yard point number, xiRepresent the row number of goods yard point i, ziRepresent goods
The line number of site i, b is goods lattice width, and h is goods lattice height；vxFor horizontal movement speed, vzFor the speed that moves vertically.
3. selection order according to claim 2 determines method, it is characterised in that in the S6：
S61：The intended particle individual extreme value corresponding with the intended particle is carried out into crossover operation, F1 is produced
Particle, the fitness of the comparison intended particle and the fitness of the F1 particle；The optimum particle of fitness is made
For second filial generation particle, second filial generation population is produced；The second filial generation population is by least one second filial generation grain
Son is constituted；
S62, the second filial generation particle and the local extremum are carried out crossover operation, produce F3 particle, compare institute
State the fitness of second filial generation particle and the fitness of the F3 particle；The optimum particle of fitness is sub as the 4th
For particle, the 4th filial generation population is produced；The 4th filial generation population is made up of described at least one the 4th filial generation particles；
S63, the 4th filial generation particle is carried out mutation operation, produce the 5th filial generation particle, comparison the 4th filial generation particle
Fitness and the 5th filial generation particle fitness；Using the optimum population of fitness as the 6th filial generation particle, produce
6th filial generation population；The 6th filial generation population is made up of described at least one the 6th filial generation particles；
It is described using the filial generation population as next iteration parent population, including：By the 6th filial generation population
As the parent population of next iteration.
4. selection order according to claim 3 determines method, it is characterised in that in the S61：
The intended particle individual extreme value corresponding with the intended particle is intersected using partial mapped crossover operator
Operation, produces F1 particle.
5. selection order according to claim 3 determines method, it is characterised in that in the S62：
The second filial generation particle is carried out into crossover operation with the local extremum using partial mapped crossover operator, the 3rd is produced
Filial generation particle.
6. selection order according to claim 3 determines method, it is characterised in that in the S63：
The 4th filial generation particle is carried out into mutation operation using sequence variation mode is changed, the 5th filial generation particle is produced.
7. a kind of selection order determining device based on genetic idea particle cluster algorithm, it is characterised in that the selection order is really
Determining device includes setting up unit, signal generating unit, computing unit, comparing unit, cross and variation unit, replacement unit and judges single
Unit：
It is described to set up unit, for according to optimization aim, setting up the object function of selection sequencing problem；
The signal generating unit, for initializing to object function using particle cluster algorithm, generates parent population；
The computing unit, for calculating the parent population in each particle fitness；
The comparing unit, for the fitness and the individuality of the intended particle of intended particle in relatively more described parent population
The fitness of optimal location, using fitness optimal value as the intended particle individual extreme value；The intended particle is described
Any one particle in parent population；
The comparing unit is additionally operable to the fitness of each particle and the parent population in parent population described in comparison
The fitness of colony's optimal location, using fitness optimal value as the parent population local extremum；
The cross and variation unit, for according to the individual extreme value and the local extremum, carrying out to the parent population
Cross and variation is operated, and generates filial generation population；
The replacement unit, for using the filial generation population as next iteration parent population；
The judging unit, for judging whether iterationses reach default maximum iteration time, if so, exports selection order
The result of problem；If otherwise triggering the computing unit.
8. selection order determining device according to claim 7, it is characterised in that the object function is specially：
Wherein, T represents the time that selection order spends, and n represents goods yard point number, xiRepresent the row number of goods yard point i, ziRepresent goods
The line number of site i, b is goods lattice width, and h is goods lattice height；vxFor horizontal movement speed, vzFor the speed that moves vertically.
9. selection order determining device according to claim 8, it is characterised in that the cross and variation unit includes intersecting
Subelement and variation subelement：
The intersection subelement, for carrying out intersecting behaviour the intended particle individual extreme value corresponding with the intended particle
Make, produce F1 particle, the fitness of the comparison intended particle and the fitness of the F1 particle；To adapt to
The optimum particle of degree produces second filial generation population as second filial generation particle；The second filial generation population by it is described at least
One second filial generation particle is constituted；
The intersection subelement is additionally operable to for the second filial generation particle and the local extremum to carry out crossover operation, produces the 3rd
Filial generation particle, the fitness of the comparison second filial generation particle and the fitness of the F3 particle；Fitness is optimum
Particle as the 4th filial generation particle, produce the 4th filial generation population；The 4th filial generation population is by described at least one
Four filial generation particles are constituted；
The variation subelement, for the 4th filial generation particle to be carried out into mutation operation, produces the 5th filial generation particle, compares institute
State the fitness of the 4th filial generation particle and the fitness of the 5th filial generation particle；Using the optimum population of fitness as the 6th
Filial generation particle, produces the 6th filial generation population；The 6th filial generation population is made up of described at least one the 6th filial generation particles；
The replacement unit specifically for using the 6th filial generation population as next iteration parent population.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710012887.XA CN106682788A (en) | 2017-01-09 | 2017-01-09 | Method and device for selection sequence determination |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710012887.XA CN106682788A (en) | 2017-01-09 | 2017-01-09 | Method and device for selection sequence determination |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106682788A true CN106682788A (en) | 2017-05-17 |
Family
ID=58850266
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710012887.XA Pending CN106682788A (en) | 2017-01-09 | 2017-01-09 | Method and device for selection sequence determination |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106682788A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019120158A1 (en) * | 2017-12-22 | 2019-06-27 | 菜鸟智能物流控股有限公司 | Item picking method, and related apparatus |
| CN113269426A (en) * | 2021-05-18 | 2021-08-17 | 哈尔滨工程大学 | Dormitory distribution method based on heuristic improved particle swarm optimization |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101807900A (en) * | 2010-03-10 | 2010-08-18 | 北京航空航天大学 | Particle filter technology based on parallel genetic resampling |
| US20120130929A1 (en) * | 2010-11-24 | 2012-05-24 | International Business Machines Corporation | Controlling quarantining and biasing in cataclysms for optimization simulations |
| CN105469145A (en) * | 2016-01-15 | 2016-04-06 | 清华大学 | Intelligent test paper method based on genetic particle swarm optimization algorithm |
| CN105574231A (en) * | 2015-11-27 | 2016-05-11 | 上海交通大学 | Storage battery surplus capacity detection method |
-
2017
- 2017-01-09 CN CN201710012887.XA patent/CN106682788A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101807900A (en) * | 2010-03-10 | 2010-08-18 | 北京航空航天大学 | Particle filter technology based on parallel genetic resampling |
| US20120130929A1 (en) * | 2010-11-24 | 2012-05-24 | International Business Machines Corporation | Controlling quarantining and biasing in cataclysms for optimization simulations |
| CN105574231A (en) * | 2015-11-27 | 2016-05-11 | 上海交通大学 | Storage battery surplus capacity detection method |
| CN105469145A (en) * | 2016-01-15 | 2016-04-06 | 清华大学 | Intelligent test paper method based on genetic particle swarm optimization algorithm |
Non-Patent Citations (3)
| Title |
|---|
| KAO YI-TUNG等: "A hybrid genetic algorithm and particle swarm optimization for multimodal functions", 《APPLIED SOFT COMPUTING》 * |
| 於世为等: "基于粒子群遗传的混合优化算法", 《系统工程与电子技 术》 * |
| 黄慧: "基于改进粒子群算法的车间作业排序的优化及仿真研究", 《中国优秀硕士学位论文全文数据库工程科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019120158A1 (en) * | 2017-12-22 | 2019-06-27 | 菜鸟智能物流控股有限公司 | Item picking method, and related apparatus |
| CN113269426A (en) * | 2021-05-18 | 2021-08-17 | 哈尔滨工程大学 | Dormitory distribution method based on heuristic improved particle swarm optimization |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shi et al. | A modified particle swarm optimizer | |
| Williams et al. | The effects of design changes and delays on project costs | |
| CN102653801B (en) | Bell-less blast furnace top burden distribution control method established on basis of genetic algorithm | |
| Mahapatra et al. | A novel ameliorated Harris hawk optimizer for solving complex engineering optimization problems | |
| DG et al. | Differential evolution algorithm with application to optimal operation of multipurpose reservoir | |
| CN106682788A (en) | Method and device for selection sequence determination | |
| CN108074004A (en) | A kind of GIS-Geographic Information System short-term load forecasting method based on gridding method | |
| CN108803332A (en) | Based on the paths planning method for improving biogeography | |
| CN101989336A (en) | Power transmission and transformation project schedule management system and method for power system | |
| CN104504627A (en) | Test paper automatic composing method utilizing genetic algorithm | |
| CN108171268A (en) | A kind of image processing method and electronic equipment | |
| CN112562078A (en) | Three-dimensional geological analysis prediction model construction method | |
| CN102508971B (en) | Method for establishing product function model in concept design stage | |
| CN113688488B (en) | Power grid line planning method based on improved artificial fish swarm algorithm | |
| Jiang et al. | An improved adaptive genetic algorithm for mobile robot path planning analogous to the ordered clustered TSP | |
| CN103544109A (en) | Novel combined test case generation method | |
| CN110298517A (en) | A kind of logistics transportation dispatching method, device and equipment based on parallel computation | |
| Ritwik et al. | A genetic algorithm-based approach for optimization of scheduling in job shop environment | |
| CN110502505A (en) | A kind of data migration method and device | |
| Shahangian et al. | A review of the conceptual framework of the interactive cycle and modeling process used in urban water management | |
| CN108460074A (en) | Multiple row based on BloomFilter indexes establishment and application method in row deposit data library | |
| CN106934453A (en) | A kind of determination method of cubic System Material parent phase and sub alternate orientation relationship | |
| Honiden | Tree structure modeling and genetic algorithm-based approach to unequal-area facility layout problem | |
| Chompu-Inwai et al. | Optimal cost driver selection in activity-based costing using Shuffled frog leaping algorithm | |
| CN107608913A (en) | A kind of CPU addressing methods, device and its CPU addressing equipment used |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170517 |